Conventionally, as a method of generating an image with wider dynamic range (referred to as WD image below), a method has been known in which a first pixel and a second pixel having different sensitivities are provided on a pixel array such as a CMOS image sensor to synthesize a first image and a second image including respective outputs from the first pixel and the second pixel. Specifically, a synthesis rate is calculated on the basis of at least one of signal levels of the first image and the second image, a moving body is detected on the basis of a difference between the first image and the second image, and a synthesis rate of a region where the moving body is detected is adjusted. Accordingly, deterioration in quality of an image caused by artifacts of a moving subject, noise at the boundary of synthesized images, and the like can be prevented.
Here, as a method of providing the pixels having different sensitivities, for example there are a method of providing a pixel having a long exposure time and a pixel having a short exposure time and a method of providing a pixel having a wide light receiving area and a narrow light receiving area (for example, refer to Patent Document 1).